Design and Implementation of High Speed CRC Generators

Transcription

1 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India Design and Implementation of High Speed CRC Generators ChidambarakumarS 1, Thaky Ahmed 2, UbaidullahMM 3, VenketeshK 4, JSubhash 5 UG Scholar, Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India 1,2,3,4 Associate Professor, Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India 5 ABSTRACT: High speed data transmission is a necessary requirement of current smart generation The networking environment is in an urge to develop extremely fast wireless communication with null occurrence of error Cyclic redundancy check (CRC) is a promising solution for error detection in data transmission The high speed data transmission with high speed error checking using parallel CRC generation is presented The proposed scheme reports parallel CRC architecture composed of 32 bits based on F matrix with a polynomial generator polynomial of order n The improved efficiency of the proposed technique verified using Xilinx ISE Simulator KEYWORDS: Cyclic Redundancy Check, Parallel CRC calculation, Linear Feedback Shift Register, LFSR, F matrix I INTRODUCTION Cyclic redundancy check is widely used in various application areas like data communication, data storage, data compression for reducing errors in data transmission The conventional hardware implementation of CRC calculations is based on the linear feedback shift registers (LFSRs) where the data is checked in a serial manner with minimum throughput Parallel CRC computation can be able to rectify these limitations to some extent such that the throughput of the 32-bit parallel calculation of CRC-32 can achieve several gigabits per second But this minimum performance improvement cannot satisfy the data transmission speed of Ethernet networks which is almost used in a huge manner Hence an alternative approaching of processing multiple bits in parallel is suggested Variants of CRCs are used in applications like CRC-16 BISYNC protocols, CRC-32 in Ethernet frame for error detection, CRC-8 in ATM, CRC-CCITT in X-25 protocol, disc storage, SDLC, and XMODEMAlbertengo et al developed z-transform based architecture for parallel CRC where synthesizable Verilog code generation is impossible Braun et al proposed a technique suitable for FPGA implementation with analytical proof of relatively high complexity Another approach based on Galois field has been proposed by Shieh et al Campobello et al suggested 32-bit parallel processing using pre-calculated F matrix whose performance will be affected by a polynomial change This paper deals with an architecture where 32-bit parallel processing is performed by means of F matrix which can be able to reduce the number of cycles of CRC to 50%This paper is organized as follows The conventional serial CRC generation technique based on LFSR is discussed in section II The parallel CRC generation scheme based on F matrix for 32 bits is presented in section III and IV respectively Simulation results are provided and discussed in section V The enhanced performance of proposed technique is concluded in section VI II SERIAL CRC Linear feedback shift registers (LFSR) are commonly used for generating serial CRC by performing a simple computation of binary divisions It can be performed by a sequence of subtractions and shifting operations In modulo -2 arithmetic, the addition and subtraction are equivalent to bitwise XORs ( ) and multiplication is equivalent to AND ( ) The basic architecture of LFSR used for serial CRC calculation is shown in Fig 1 Copyright to IJIRSET wwwijirsetcom 219

2 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India Fig 1: Basic LFSR Architecture As shown in above Fig d is serial input data, X represents present state (generated CRC), X represents next state and p is generator polynomial The operation of basic LFSR architecture is expressed in terms of following equations X 0 ' = (P 0 X m 1 ) d (1) X i ' = (P 0 X m 1 ) X i 1 The generator polynomial for CRC-32 is as follows G(x) = x 32 + x 26 + x 23 + x 22 + x 16 +x 12 +x 11 +x 10 +x 8 + x 7 + x 5 + x 4 + x 2 + x 1 + x 0 ; We can extract the coefficients of G(x) and represent it in binary form as P = {P32, P31, P0} P = { } Frame Check Sequence (FCS) will be generated after k+m cycle, where k denotes the number of bits of data and m denotes the order of generator polynomial In case of 32 bits serial CRC, the serial CRC generation will happen after 64 cycles when the order of generator polynomial is 32 III PARALLEL CRC There multiple techniques used for parallel generation of CRC is labeled below 1 A Table-Based Algorithm for Pipelined CRC Calculation 2 Fast CRC Updation 3 F matrix based parallel CRC generation 4 Algorithm for Unfolding, Re-timing and Pipelining LUT based architecture provides minimum memory, whereas the high pipelining table based architecture has multiple inputs such as LUT1, LUT2 and LUT3 LUT3 contains CRC values for the input followed by 12 bytes of zeros, 8 bytes of LUT2, and 4 bytes of LUT4 Higher throughput has been achieved through this algorithm But its performance is limited by pre-calculating CRC and storing it in LUT, because LUT must be changed with respect to polynomial change Pipelining algorithm can be used to reduce the critical path with the integration of delay element Copyright to IJIRSET wwwijirsetcom 220

3 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India Fig 2: LUT based architecture Parallel processing used to increase the throughput by producing the number of output at the same time Retiming is used to increase the clock rate of the circuit by reducing the critical path computation time The fast CRC update technique involves the calculation of CRC for the bits which undergoes any change instead of calculating CRC each time for all the data bits There are different approaches to generate the parallel CRC having advantages and disadvantages for each technique Table based architecture required pre-calculated LUT, so, it will not be used for generalized CRC, fast CRC update technique required buffer to store the old CRC and data In unfolding architecture increases the no of iteration bound The F matrix based architecture more simple and low complex A ALGORITHM FOR F MATRIX BASED ARCHITECTURE Algorithm and Parallel architecture for CRC generation based on F matrix is discussed in this section As shown in fig 2 it is basic algorithm for F matrix based parallel CRC generationparallel data input and each element of F matrix, which is generated from given generator polynomial is ANDed, result of that will XORing with present state of CRC checksum The result generated after (k+ m) /w cycle Fig 3: Algorithms for F matrix based architecture B F MATRIX GENERATION F matrix is generated from generator polynomialwhere, {p 0 p m-1 } is generator polynomialfor example, the generator polynomial for CRC4 is {1, 0, 0, 1, 1} and w bits are parallelly processed Copyright to IJIRSET wwwijirsetcom 221

4 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India F = Pm Pm F= Pm Pm P C PARALLEL ARCHITECTURE Parallel architecture based on F matrix illustrated in fig 3 As shown in fig 1, d is data that is parallel processed (ie 32bit), X' is next state, X is current state (generated CRC), F(i)(j) is the i th row and j th column of F matrix If X = [x m-1 x 1 x 0 ] T is utilized to denote the state of the shift registers, in linear system theory, the state equation for LFSRs can be expressed in modular-2 arithmetic as follow X i '= (P 0 X m 1 ) X i+1 (2) Where, X(i) represents the i th state of the registers, X(i + 1) denotes the (i + 1) th state of the registers, d denotes the one-bit shift-in serial input F is an m x m matrix and G is a 1 x m matrix Finally, equation (3) can be written in matrix form as G = [ ] T (3) X ' = F W X d (4) Equation (4) can be expanded as, X 3 '= X 2 X 1 X 0 d 3 X 2 '= X 3 X 2 d 2 X 1 '= X 3 X 2 X 1 d 1 X 0 '= X 3 X 2 X 1 X 0 d 0 (6) The dividend is divided into three 4-bit fields, acting as the parallel input vectors D(0), D(1), D(2), respectively The initial state is X(0) = [ ] T X(4) = F 4 X(0) D(0) X(12) = F 4 X(8) D(2) (7) calculation of the next state X by matrix F w, current state X and parallel input D, make the 32-bit parallel input vector suitable for any length of messages besides the multiple of a sequence of 32-bit parallel calculation, the final remaining number of bits of the message could be 8; 16,or 24 For all these situations, an additional parallel calculation w = 8; 16; 24 is needed by choosing the corresponding F w If the length of the message is Copyright to IJIRSET wwwijirsetcom 222

5 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India not the multiple of the number of parallel processing bits w = 4 ie data bit is Then last two more bits D(3) need to be calculated after getting X (12) For calculating the extra bits if the data message length is not the multiple of w, the number of parallel processing bits It is worth to notice that in CRC-32 algorithm, the initial state of the shift registers is preset to all `1's Therefore, X(0) = 0xFFFF However, the initial state X (0) does not affect the correctness of the design For better understanding, the initial state X (0) is still set to 0x0000 when the circuit is implemented IV PROPOSED DESIGN In proposed architecture w= 32 bits are parallelly processed and order of generator polynomial is m= 16 As discussed in section 3, if 32 bits are processed parallelly then CRC-32 will be generated after (k m)/w cycles If we increase number of bits to be processed parallelly, number of cycles required to calculate CRC can be reduced Proposed architecture can be realized by below equation Where, Xtemp= F W D(0to15) D(16to32) X' = F W X Xtemp (11) D (0 to 15) =first 16 bits of parallel data input D (0 to 32) = next 16 bits of parallel data input X =next state,x=present state Fig 4: Circuit diagram of parallel calculation of CRC-32 In proposed architecture d i is the parallel input and F(i)(j) is the element of F 16 matrix located at i th row and j th column As shown in figure 3 input data bits d 0 d 15 anded with each row of F W matrix and result will be xored individually with d 16, d 17 d 32 Then each xored result is then xored with the X' (i) term of CRC-32 Finally X will be the CRC generated after (k m)/w cycle, where w=3 A SYNTHESIS REPORT V RESULT The proposed architecture is synthesized in Xilinx-92i and simulated in Xilinx ISE Simulator, which required half cycle then the previous 32-bit design In our programming in Verilog by specifying only generator polynomial, it directly gives F matrix useful for parallel CRC generation that is not available in previous methods Hardware utilization is compared in table I for different approaches for different parameter like LUT, CPD and cycle Copyright to IJIRSET wwwijirsetcom 223

6 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India fig 7: Synthesis report for parallel CRC-32 Fig 8: Total delay of parallel CRC-32 CRC W parallel bits Cloc k cycle LUTs CPD CRC32 w=16bit[1 ] CRC32 w=32bit[8 ] CRC32 w=32bit(proposed) ns ns ns Fig 9: Comparison table for different algorithms of CRC Copyright to IJIRSET wwwijirsetcom 224

7 Department of ECE, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India B SIMULATION From the table it is observe that, the architecture proposed by require 17 clock cycle to generate CRC as per equation (k+ m)/w and for proposed architecture it required only 9 cycle, CPD for proposed architecture is less than the architecture, only the disadvantage for proposed architecture is the no of LUT get increased, so area also get increase VICONCLUSION 32-bit parallel architecture required 17 ((k + m)/w) clock cycles for 32-byte data Proposed design (32bit) required only 9 cycles to generate CRC with same order of generator polynomial So, it drastically reduces computation time to 50% and same time increases the throughput Pre-calculation of F matrix is not required in proposed architecture Hence, this is compact and easy method for fast CRC generation REFERENCES [1] Campobello, G; Patane, G; Russo, M; "Parallel CRC realization," Computers, IEEE Transactions on, vol52, no10, pp , Oct2003 [2] MDShieh et al, A Systematic Approach for Parallel CRC Computations, Journal of Information Science and Engineering, May 2001 [3] Braun, F; Waldvogel, M; "Fast incremental CRC updates for IP over ATM networks," High Performance Switching and Routing,2001 IEEE Workshop on, vol, no, pp48-52, 2001 [4] Weidong Lu and Stephan Wong, A Fast CRC Update Implementation, IEEE Workshop on High Performance Switching and Routing, pp , Oct 2003 [5] RameshwarMurade, MD Manan Mujahid, MAM Sabir, The design and implementation of a Programmable CRC computation circuit architecture using FPGA 2013 [6] SR Ruckmani, P Anbalagan, High Speed cyclic Redundancy Check for USB Reasearch Scholar, Department of Electrical Engineering, Coimbatore Institute of Technology, Coimbatore , DSP Journal, Volume 6, Issue 1, September, 2006 [7] Yan Sun; Min Sik Kim; "A Pipelined CRC Calculation Using Lookup Tables," Consumer Communications and NetworkingConference (CCNC), th IEEE, vol, no, pp1-2, 9-12 Jan2010 [8] Sprachmann, M; "Automatic generation of parallel CRC circuits," Design & Test of Computers, IEEE, vol18, no3, pp , May 2001 Copyright to IJIRSET wwwijirsetcom 225